Data Diving
Published in

Data Diving

Marine Species Trends in the Caribbean Sea

Ok, here’s a million dollar question: is the Caribbean really dying? Or, more specifically, are marine species found on Caribbean reefs becoming less abundant?

To answer this question we’ll need to pick our weapons of analysis, namely…

  1. An annual time series plot.
  2. Linear regression with weights.

Surprisingly, these tried and true methods give us plenty of insight into the data (we’ll find an excuse to break out neural nets another day). In the meantime, we have four plots below, each of which shows a different marine organism’s average annual count over the past 22 years. So, let’s jump right in.

Analysis of Hard and Soft Corals

The first of our charts depicts the centerpiece, the make-or-break of any coral reef: hard corals. Hard corals, or reef-building corals, are fantastically intricate living rocks. They provide both food and shelter to a wide range of marine organisms. Long story short, without reef-building corals, reefs would not exist.

Figure 1: Average annual % cover for hard coral on Caribbean reefs. Data source is Reef Check.

Now the amount of hard coral is often measured by percent coverage, that is, from a birds-eye-view, the percent of an area that is covered. In Figure 1 (left), we can see that the weighted regression line has a slope of -0.3; this corresponds to a 0.3 decrease in percent coral coverage every year.

0.3%? That’s it? Yes we’ve lost roughly 7.2% of our coral coverage in the past 22 years. But remember this is percent coverage; using the fitted values, the total percent change is -28%. Moreover, prior research suggests that since the 1950’s, we’ve lost around 50% of Caribbean hard corals.

Dying: 1.

Not-Dying: 0.

For our second case, let’s look at soft corals. Soft corals are the Robin to a hard coral’s Batman. Rooted at the base similar to a tree, they grow upwards in search of sunlight and can often be seen flowing in ocean currents. While they have less structural importance than their harder counterparts, soft corals are still an integral part of reef food chains.

Figure 2: Average annual % cover for soft coral on Caribbean reefs. Data source is Reef Check.

As shown in Figure 2, soft corals show an even stronger negative trend relative to their initial value. They too have decreased by about 7.2%, but because they started with only 11.6% total cover, their percent change is -57.4%. In 2019, the average diver would see a striking 1.7% cover for soft corals.

Despite this dismal trend, it’s interesting to note that after almost every decrease in percent cover, soft corals appear to rebound. Once again we have evidence that nature can be resilient when given the opportunity.

Reasons for coral mortality are well-documented and have been for the past 15 years. Despite a long list of culprits, the most cataclysmic cause of coral death is the infamous coral bleaching, which occurs when water temperatures become too hot or cold. There were around five main bleaching events in the Caribbean that overlap with our time series, but I would like to highlight three that occurred in 1997, 2005, and 2015. Let me know if you see any relationships.

Analysis on Fish

The above information is not new. However, relatively little time series analysis has been done on things that move, for example, fish.

So, on to the third of four figures. Here we will be looking at the prolific yet uniquely colorful butterfly fish. Fun fact of the day: butterfly fish mate for life. A less fun fact (but still reasonably fun) is butterfly fish are omnivorous but rely on coral for a large portion of their diets.

Figure 3: Average annual butterfly fish counts on Caribbean reefs. Data source is Reef Check.

As shown in Figure 3 to the left, butterfly fish counts have declined precipitously over the past 22 years. Moreover, there is a striking resemblance to the trends exhibited by hard coral, lagged by roughly a year. For instance, the plunge in coral from 2006–2011 is almost identical to the plunge in butterfly fish from 2007–2012.

Using a weighted regression line to summarize, over the past 22 years the average number of butterfly fish per dive has decreased by a count of 10, equating to a 52.5% decline.

Dying: 3.

Not-Dying: still 0.

It’s not looking good

On to the final figure. In this graph we will be discussing the aptly named parrot fish. Arguably the ugliest beautiful fish, the parrot fish is brightly colored from beak to tail. Yes, it has a beak.

Figure 4: Average annual parrot fish counts on Caribbean reefs. Data source is Reef Check.

Parrot fish use this beak to break off chunks of hard coral and because they adhere to a strict coral diet, one would expect their numbers to be closely linked to hard coral percent cover.

According to Figure 4, that theory has some supporting evidence. In 2005, the year of arguably the most catastrophic bleaching event in recent Caribbean history, we see a massive drop-off in fish counts. However, the symmetry between butterfly fish and hard coral is surprisingly not observed with parrot fish.

Now, why the spike in 2018? Great question. After looking through the literature, there was no conclusive evidence.

However, despite our lack of understanding of the drivers of these trends, we can still observe and summarize them: parrot fish have decreased by 48% over the past 22 years with a year over year change of -0.57 in counts per dive.

A side note, parrot fish showed one of the strongest negative trends out of all fish observed.


Hopefully this analysis was able to shed some light onto our initial question of “is the Caribbean really dying?” While these four species (or grouping of species in the case of hard and soft corals) are by no means a representative sample of the Caribbean, they certainly can be used as proxies for reef health. And, it’s clear that hard corals, soft corals, butterfly fish, and parrot fish are all in decline.

Personally, what makes these problems so interesting (other than sharks are cool) is the fact that a large portion of society has ignored these findings. Is it charity burnout? Maybe. But I think observing these trends is almost pointless; we already know this stuff. Instead, we need to build tools that quantify and forecast ocean damage, ideally in dollars.

That’s my humble opinion. If you have thoughts or ideas, please leave a comment or reach out.


  • Animals, A. (2019, November 08). Butterfly Fish. Retrieved August 14, 2020, from
  • Buddemeier, R. W. (2003). Coral Reef Decline in the Caribbean. Science, 302(5644), 391c-393. doi:10.1126/science.302.5644.391c
  • Caribbean coral reef decline began in 1950s and 1960s from local human activities. (2020, April 22). Retrieved August 14, 2020, from
  • Crcp. (2016, April 19). NOAA’s Coral Reef Conservation Program (CRCP) — Coral Facts. Retrieved August 15, 2020, from
  • Muñiz-Castillo, A. I., Rivera-Sosa, A., Chollett, I., Eakin, C. M., Andrade-Gómez, L., Mcfield, M., & Arias-González, J. E. (2019). Three decades of heat stress exposure in Caribbean coral reefs: A new regional delineation to enhance conservation. Scientific Reports, 9(1). doi:10.1038/s41598–019–47307–0

The data were collected by Reef Check, a coral conservation non-profit that trains volunteer divers to collect marine data. With 1576 unique entries for the Caribbean ranging from 1997–05–24 to 2019–08–24, there were plenty of data points to conduct a TS analysis. However, due to variations in the dive sites, specifically location and time, annual means for the entire Caribbean were calculated. Finally, the least squares regression line was weighted to account for the number of samples each year; early and modern years had less dives than middle years.

Here is the code.

Note: these are my findings. If you would like to contact me, leave a message here. All criticisms are welcome.



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Michael Berk

Michael Berk

I’m a Data Scientist writing 52 posts that bring academic research to DS industry.